With the rapid development of social media, social factors have become an important factor to affect the group decision making process and its result. For the group decision-making problems where decision information is given by incomplete preference relations, this paper proposes effective solutions. Firstly, we propose a repairing method based on the similarity degree between any two decision makers and social network distances. Then, we propose the social network influence based group interactive consensus model. The model not only considers the social adjacency relation between group decision makers, but also can save the original judgment information given by decision makers on a large extent. Finally, an example of logistics enterprises choosing storage warehouse is given to illustrate the feasibility and advantages of the proposed approach.